unyer
20/07/2026

Online sales and marketplaces: The new frontier of distribution control

Abstract

 

The distribution battle has moved online.

 

The competition law framework governing online sales has undergone a seismic shift. The Digital Markets Act (DMA[i]), which came into force on 4 November 2024, has fundamentally reshaped how suppliers can control digital distribution channels. Simultaneously, the Platform to Business Regulation (P2B)[ii] and sectoral enforcement across EU Member States – including high-profile cases before the CJEU and national courts – have exposed the tensions between brand preservation and competition principles.

 

For suppliers in high-image sectors, such as luxury, cosmetics, high-tech, the central question has shifted. Selective distribution itself remains lawful. What has become increasingly perilous is the contractual architecture surrounding digital channels: marketplace restrictions, approval mechanisms, and pricing controls that once seemed defensible now invite intense scrutiny.

 

This article dissects the current legal landscape governing online sales in selective distribution networks, drawing on recent CJEU case law, DMA enforcement signals, and French jurisprudence. It argues that the days of reflexive marketplace bans are over. What survives – and what thrives – is proportionate, objectively justified contractual control over online sales, carefully calibrated to the competitive reality of omnichannel retail.

 

It is recommended to:

 

  • distinguish clearly between restriction types
  • document the objective and proportionality analysis
  • consider market-specific rules
  • account for the DMA and national enforcement
  • be proportionate, not reflexive

 

On this frontier, claims are staked with precision – vague restrictions get jumped.

 

Selective distribution in the digital age: a tool for the marketplace era – From principle to practice

 

European competition law has long endorsed selective distribution networks. The principle is venerable: a supplier may select distributors on uniform, non-discriminatory criteria without breaching Article 101 TFEU – provided the criteria are objective and applied even-handedly (CJEC, 25 October 1977, Metro SB-Großmärkte GmbH & Co. KG v Commission of the European Communities, 26/76)[iii].

 

The presence of online sales does not invalidate this permission. A brand may legitimately require that its products be sold in a retail environment – whether physical or digital – consistent with its positioning and perceived quality. As the CJEU confirmed in Coty Germany, the supplier’s interest in preserving brand image through controlled distribution channels is a recognised legitimate objective[iv].

 

What has changed is not the principle, but its application. The rise of marketplaces, social commerce, and influencer-driven retail has forced lawyers and compliance teams to rethink what “control” actually means in a digital economy where visibility, pricing transparency, and seller reputation are fungible commodities.

 

This is the Wild West of e-commerce, and brands are learning that hired guns and blanket bans don’t survive long against a savvy adversary.

 

Marketplaces: The control problem

 

A marketplace is not a neutral conduit – it is a competitive arena where brand owners lose command over key variables: presentation, juxtaposition with rival products, customer reviews, and pricing visibility. For a cosmetics brand concerned with heritage and prestige, the difference between a direct-to-consumer website and a Zalando storefront is categorical. On Zalando, for example, the brand sits alongside budget alternatives. On its own site, it controls the narrative.

 

This is why so many selective distribution agreements now contain clauses prohibiting or restricting sales through third-party platforms. The legal question has become: How restrictive can such clauses be before they cross into hardcore territory?

 

The answer, according to established case law, depends on whether the restriction genuinely serves a legitimate purpose or is merely a pretext for partitioning markets. Pierre Fabre case remains the lodestar: an absolute prohibition on all online sales constitutes a hardcore restriction unless it can be justified by the nature of the product and the requirements of its marketing[v].

 

French courts have acknowledged limited exceptions. In Bang & Olufsen, the Cour d’appel de Paris held that a blanket online prohibition could be proportionate for high-end audio equipment, given the technical complexity and need for expert advice[vi]. But such cases are narrow. They do not herald a return to blanket bans; they recognise that context matters.

 

That’s where the outlaws get hanged: when control masquerades as protection and, actually, just partitions the territory.

 

The danger zone: When proportionality collapses (and the line is crossed)

 

A marketplace restriction tips into hardcore territory when:

 

  • It prevents passive sales to end customers in other Member States (cross-border restriction);
  • It effectively forecloses online selling altogether, even if nominally permitting a website;
  • It imposes minimum resale prices or actively polices discounts;
  • It requires pricing approval that is de facto gatekeeping;
  • It makes the digital channel so burdensome (documentation, audits, approval delays) that it becomes economically unusable.

 

The CJEU has signalled that such restrictions fall outside the scope of permissible brand protection. What matters is not ‘whether’ a brand wishes to preserve its image, but ‘how’ it goes about doing so. A quality standard for online presentation is defensible. A requirement that the distributor operate only on the brand’s own platform is not.

 

A critical nuance: the presence of one problematic clause does not automatically doom the entire selective distribution network. French courts, pragmatically, have entertained severability arguments: if the offending clause can be excised and the network remains functional and pro-competitive, the network may survive even if the clause is struck down[vii]. But this is not a safe harbour. Severability is a last resort, not a licence for sloppy drafting.

 

French case law illustrates this principle. In Garage de Bretagne v. Mercedes-Benz France, The Cour de cassation held that neither EU law nor French competition law prohibits the mere refusal by a supplier at the head of a qualitative selective distribution network to approve distributors who meet the selection criteria.

 

The Court confirmed that the network as a whole could remain valid even where the supplier’s exercise of discretion in approving distributors was contested. The mere presence of a disputed refusal-to-approve decision did not invalidate the underlying network structure[viii]. More broadly, the Paris Court of Appeal and French Cour de cassation have consistently adopted this pragmatic stance: even if certain contractual clauses are classified as hardcore restrictions, this does not rule out the possibility that the selective distribution network complies with competition rules, subject to an in-depth analysis of the practices implemented and their impact on the market. The mere presence of a black clause does not render the entire network illegal per se[ix].

 

In cases where the problematic clause can be isolated – whether by severability or narrow interpretation – the clause may simply be deemed unenforceable without invalidating the entire network.  This approach does not, however, constitute a safe harbour. Courts conduct a rigorous competitive effects analysis. As the rulings from 2019 and 2023 demonstrate[x], the lawfulness of a selective distribution network must be assessed on a case-by-case basis, depending on the actual effects of the clauses on competition. A supplier cannot draft a network carelessly and rely on severability to rescue it. The clause must genuinely be severable – not essential to the network’s operation or brand strategy – and the remaining network must demonstrably continue to serve a legitimate purpose and operate competitively.

 

The DMA factor: A new regulatory layer

 

Since November 2024, large online platforms designated as “gatekeepers” under the Digital Markets Act face obligations that ripple through the distribution ecosystem. Gatekeeper platforms must grant non-discriminatory access to third-party sellers, provide transparent algorithms, and refrain from self-preferencing. While the DMA does not directly govern selective distribution agreements, it has reshaped marketplace dynamics in ways that suppliers must consider when drafting restrictions.

 

A brand that prohibits sales through a DMA-designated gatekeeper may find the restriction economically illusory: the platform’s scale and reach mean that the restriction, while contractually binding, excludes the brand from the most high-traffic retail channel in Europe. Conversely, a brand that permits sales only on gatekeepers – to the exclusion of smaller, niche marketplaces – may run afoul of proportionality. The DMA has thus tilted the playing field in ways that rigorous proportionality analysis must now account for.

 

The Sheriff has arrived, and his name is the Digital Markets Act – the bad old days of unchecked gatekeeping are over.

 

Practical implications for brand owners and counsel

 

  1. Distinguish clearly between restriction types

 

Do not conflate a prohibition on unauthorized marketplace use with an outright ban on online sales. The former may be defensible; the latter is not. Draft clauses should explicitly state: “Distributor may sell online via its own website or through selected platforms approved in writing by the Supplier, provided (i) [quality criteria] and (ii) [service commitments].” Avoid language such as “all online sales are prohibited” or “sales through third parties are forbidden.”.

 

  1. Document the objective and proportionality analysis

 

A restrictive clause that does not articulate its justification is vulnerable. The agreement should state: “This restriction is necessary to preserve product image and ensure compliance with [specific quality or service standards] that are material to the positioning of the brand.” Be specific. Courts and authorities examine whether the stated objective is genuine and whether less restrictive means could achieve it.

 

  1. Consider market-specific rules

 

Luxury goods and high-touch categories may support narrower digital restrictions than mass-market consumer goods. A haute couture brand defending a marketplace ban is on firmer ground than an appliance manufacturer. But even here, recent cases suggest that some online channel – whether the brand’s own site or approved partners – must remain available.

 

  1. Account for the DMA and national enforcement

 

When drafting marketplace restrictions, consider whether they will withstand review under the DMA, the P2B Regulation, and the precedents of the Autorité de la concurrence (France) or equivalent agencies. A marketplace ban that was legally safe in 2020 may invite scrutiny in 2026 if it has the practical effect of partitioning markets or discriminating against smaller distributors.

 

  1. Be proportionate, not reflexive

 

The question to ask is not “Can we prohibit marketplaces?” but “What is the minimum restriction necessary to preserve our legitimate brand interests?”. If a quality approval process, pricing guidelines, or presentation standards would suffice, they are preferable to an outright ban.

 

Think of these five principles as the frontier settlers’ handbook: follow them, and you’ll stake your claim securely.

 

Conclusion: Intelligent governance, not reflexive control

 

The law governing online sales in selective distribution networks is no longer in formation – it is settled. Selective distribution is lawful. Proportionate digital restrictions are lawful. Reflexive marketplace bans, blanket online prohibitions, and clauses that have the practical effect of foreclosing digital sales are not.

 

The era of “we don’t sell online” has passed. What has emerged is an era of intelligent governance: brands that can articulate clear, objective, proportionate criteria for their digital channels, and distributors that have real (if bounded) access to meaningful sales avenues. This is not a compromise that limits brands; it is a framework that protects them from legal liability whilst enabling them to manage their distribution networks effectively.

 

For counsel, the lesson is straightforward. When a brand asks “Can we ban marketplace sales?”, the answer is no – or at least, not universally and not without risk. The better question is: “How do we structure digital distribution so that it serves our brand positioning within the bounds of competition law?” That question yields compliant, durable, and commercially intelligent answers.

 

The frontier has been mapped, the rules are written, and the winners will be those smart enough to adapt rather than resist.

 

___

[i] The main legislative texts for the Digital Markets Act (DMA) are Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on contestable and fair markets in the digital sector and the Implementing Regulation

[ii] Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services

[iii] CJEC, 25 October 1977, Metro I, 26/76

[iv] CJEU, 6 December 2017, Coty Germany GmbH v Stadtsparkasse Magdeburg, C-230/16, para 37

[v] CJEU, 13 October 2011, Pierre Fabre Dermo-Cosmétique, C-439/09, paras 45–48

[vi] CA Paris, 13 March 2014, SARL Bang & Olufsen France, n° 2013/00714

[vii] “Black clauses” & Selective distribution agreement – Unyer

[viii] Cour de cassation, 16 February 2022, No. 20-11.754

[ix] “Black clauses” & Selective distribution agreement – Unyer

[x] “Black clauses” & Selective distribution agreement – Unyer

Alban Curral
Summary of the privacy policy
Unyer

Cookies - Functions

This cookie is used to store your cookie preferences :
- _moove gdpr pop up .

Web analytics cookies

This website uses Google Analytics for statistical purposes (audience measurement cookies). They make it possible to know how many times a certain page has been consulted. We only use this information to improve the content of our website.

Please find below the cookies used :

– ga : This cookie is used to identify users of the Website via the users’ IP addresses.

– gat : This cookie is used to limit the number of simultaneous requests in order to avoid bugs

– gid : This cookie is used to identify the Website users via their IP address (which are stored 24h)

You can visit Google's privacy page.